The fourth key factor in learning is consolidation. This is the progressive automation of the circuits underlying learning. Automation transfers acquired knowledge from the conscious compartment to specialized, non-conscious circuits, freeing up mental resources for new tasks. Brain imaging during reading or mental arithmetic shows that, at the start of learning, there is massive activation of the executive control circuits associated with the prefrontal cortex. Attention is strongly mobilized, and memorized information is processed explicitly, consciously, sequentially and with effort. Gradually, this activity is reduced, while it increases in certain specialized areas of the posterior regions of the brain. In the field of reading, for example, a child's initial deciphering results in a linear increase in reading time as a function of the number of letters in a word. As reading becomes more automatic, reading time accelerates and becomes constant whatever the length of the word (between 3 and 8 letters). The child can then concentrate on the meaning of the text.
Sleep appears to be a major factor in the consolidation of learning. As early as 1924, Jenkins and Dallenbach showed that response time accelerated after a period of sleep. It wasn't until the work of Karni et al (1994), followed by the Stickgold, Ribeiro and Born teams, that cognitive and motor performance improved significantly and durably after a period of sleep, without any additional training, whereas sleep disruption selectively blocked this consolidation. The respective roles of the different sleep stages are not yet perfectly established, but it would seem that deep, slow-wave sleep enables the consolidation and generalization of knowledge (declarative memory), while REM sleep reinforces perceptual and motor learning (procedural memory).
